Browse Articles
Discover research articles across all indexed journals
miR-139-5p suppresses hepatocellular carcinoma progression by targeting SMOX to inhibit AKT-mTOR pathway and epithelial-mesenchymal transition
Randomly distributed optical fibers in translucent mortar for privacy-preserving light transmission and digital image reconstruction
Abstract Light-transmitting building materials often compromise visual privacy due to coherent light transmission. This study presents a novel composite utilizing randomly distributed optical fibers coupled with a computational image reconstruction system. A limestone-calcined clay cement (LC³) inspired matrix was designed for sustainability and material performance. An algorithmic approach assigned a random traceable fiber distribution via a bijective input-output mapping. The random fiber configuration achieves effective light diffusion, preserving physical privacy. However, using digital imaging and homography-based calibration, the network was computationally reconstructed to reverse the diffusion, recovering hidden visual information accurately. This demonstrates a dual functionality: architectural privacy combined with selective digital transparency. Geometric robustness tests confirmed a stable operational envelope (estimated error of 2.9%) across viewing distances of 30–110 cm and camera rotation angles up to ± 35º (pitch and yaw), establishing these fiber-instrumented cementitious composites as hybrid physical-digital materials for smart infrastructure applications.
Clustering lung function and symptom profiles for asthma risk stratification
Abstract Asthma is a heterogeneous condition often studied through wheeze alone, yet the interplay between lung function and reported symptoms remains underexplored. To capture this heterogeneity, we applied Bayesian Profile Regression to data from school-age children in two prospective birth cohorts, integrating airway hyperresponsiveness, lung function, bronchodilator reversibility, allergic sensitisation, reported symptoms, and physician diagnosis. In the Manchester Allergy and Asthma Study (discovery cohort), five reproducible clusters were identified: HA-LLF (high asthma-low lung function), HA-NLF (high asthma normal lung function), LA-RLF (low asthma-reduced lung function), LA-NLF (low asthma normal lung function), and MA-NLF (moderate asthma normal lung function). The HA-LLF and HA-NLF clusters had very high asthma prevalence (80–100%), but differed markedly in lung function, airway responsiveness, bronchodilator reversibility, sensitisation, and symptom burden. The LA-RLF and LA-NLF clusters with low asthma prevalence (< 5%) displayed contrasting lung function profiles, while MA-NLF (~ 50% asthma prevalence) was largely defined by prominent symptoms such as chest tightness and shortness of breath. These subtypes were replicated in an independent cohort, Isle of Wight. Our findings demonstrate that integrating physiological, immunological, and symptom-based measures yields clinically meaningful asthma subtypes beyond wheeze-based definitions and may support more precise disease classification.
Comparison of the frailty phenotype and frailty index for identifying vulnerable companion dogs
Probiotics intervention reduces oxidative stress–driven myocardial injury
Creative and depressive profiles shape divergent thinking in emotion regulation idea generation
Engineering rhodium encapsulated indium doped fullerene for NH3, NO, and NO2 sensing
Abstract Nitrogenous gas pollutants, such as NH 3 , NO, and NO 2 , contribute significantly to environmental degradation, driving water pollution, biodiversity loss, and impaired air quality while posing critical risks to human health. Despite advancements in gas sensing technologies, materials with enhanced sensitivity and selectivity remain crucial for reliable pollutant detection. In this work, we investigate the gas sensing properties of rhodium-encapsulated indium-doped carbon-based fullerene (In-Rh@C 60 ) via DFT/PW6B95-D3/GenECP and ωB97X-D/LANL2DZ computational methods. The energy gap values were found to range from 0.705 to 1.537 eV and 3.980 to 5.166 eV for the respective methods. Upon adsorption of NH 3 , NO, and NO 2 , the energy gap decreases, indicating enhanced sensitivity. The observed chemisorption phenomena exhibit adsorption energies between -13.49 and -8.397 eV. Notably, adsorption of NO at the O-site and N-site leads to the most pronounced energy gap reductions (-0.515 eV and -0.389 eV). The green regions observed in the 3D non-covalent interaction plots signify strong van der Waals interactions, contributing to the stability of the adsorbent-adsorbate systems. This tailored system demonstrates suitability as an adsorbent material for gas pollutants, with adsorption site specificity taken into account. The findings suggest that the modified In-Rh@C 60 system holds potential as an adsorbent material for integration into sensor devices aimed at detecting NH 3 , NO, and NO 2 gas pollutants.
Correction: Innovative ZnO-W18O49 nanocomposites and ZnWO4 nanostructures for water treatment
Pre-activation timing determines influenza severity and viral pathogenicity via STING Inhibition
Harnessing hyperspectral imaging and machine learning techniques for accurate discrimination of peanut plants and weeds
Range-wide assessment of habitat suitability for jaguars using multiscale species distribution modelling
Abstract Jaguars ( Panthera onca ) are highly sensitive to persecution, habitat loss, and fragmentation, making the identification of suitable habitat critical for conservation planning. Using GPS telemetry data from 172 individuals across eight countries – the largest jaguar dataset to date – we developed multiscale Resource Selection Functions (RSFs) incorporating 15 environmental covariates to model habitat suitability across the species’ historic range. Jaguars selected productive habitats near water and strongly avoided human-modified landscapes, including areas with high human population density and livestock presence. The resulting habitat suitability surface showed strong predictive performance (AUC = 0.88; Boyce Index = 0.91) and correlated with known density estimates and distribution models. Jaguar Conservation Units (JCUs) and Protected Areas (PAs) contained 68.7% and 53.9% of predicted suitable habitat, respectively, while occupying only a third of the range. Non-designated lands, though comprising just 4% of the range, held nearly 10% of total suitability. The Amazon and Mayan Forests were identified as core strongholds, while ecoregion-based modelling revealed additional areas of high suitability in the Pantanal, Gran Chaco, Cerrado, and coastal Mexico. While Brazil encompassed the largest extent of highly suitable habitat, countries such as Paraguay, Argentina, and the United States gained conservation relevance under the ecoregion-stratified scenario.
Late posttonsillectomy haemorrhage, differences in unilateral versus bilateral tonsillectomy, a retrospective epidemiological multicentric study
Impact of lactation stage, maternal age, and parity on mammary epithelial cell populations in fresh milk
Recombinant Bacillus subtilis spores expressing SARS-CoV-2 spike protein induced humoral, mucosal, and cellular immunity in mice
The Nature Podcast highlights of 2025
DDRN: DETR with dual refinement networks for autonomous vehicle object detection
Federated learning with LSTM and error correcting codes for secure and private identification of IoT devices
A reinforcement learning algorithm to optimize resource utilization in combat casualty care
Abstract Fluid resuscitation immediately following a hemorrhagic injury improves clinical outcome. However, future military conflicts are expected to result in mass-casualty incidents with limited availability of fluid-resuscitation resources. Here, we developed and assessed the performance of a reinforcement learning AI method that optimized clinical outcome and fluid allocation under constrained resources. We generated a large cohort of synthetic trauma casualties using a validated cardio-respiratory computational model, simulating vital-sign time-series data for realistic battlefield scenarios involving hemorrhage, tourniquet application, and fluid resuscitation. For each casualty, we simulated three intervention options—infusion or no infusion of one fluid unit every 30 min over 90 min—and assessed whether the interventions restored the vital signs to “healthy” levels. Using these data, we trained the reinforcement learning model to predict the optimal sequence of interventions that maximized the number of casualties restored while minimizing fluid utilization. Using independent simulated data, we found that the AI model was twice as efficient and restored more than twice as many casualties as the current standard of care across varying numbers of casualties and resource limitations. These results highlight the model’s potential to enable personalized interventions, enhance treatment efficiency, and support automated medical decision-making in resource-constrained environments.